Year: 2020
Type: dissertation
Abstract: Time to disease progression or death (PFS) is of main interest in oncology clinical trials. The data is commonly analyzed using a Cox model. Early discontinuation gives uncertainty in the estimated hazard ratio (HR). Censoring is generally used to handle discontinuations. However, if discontinuations are related to the patients’ prognosis, the HR may be biased. By simulating PFS data, this bias was investigated. The direction and size of the bias is dependent on the proportion of censored patients in each treatment arm. Our results show that supplementary analyses with different strategies to handle discontinuation should be used to compare cancer treatments.
Author: Andrea Mattsson
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3086292605